We propose co complete the neurodevelopmental assessment at age 2 years of the 922 survivors of a unique cohort of 1,105 births under 2000g enrolled between 1984 and 1987 in a multi-center. geographically based study of the causes and consequences of neonatal brain hemorrhage. Current funding will only allow us to complete approximately 54% of the assessments, and a very partial analysis of those data that have been collected. Hemorrhage was ascertained through systematic, timed cranial ultrasound screening of all enrolled infants at 4 hours, 24 hours, and 7 days postnatally. Data were collected for all infants on pregnancy complications, labor and delivery abnormalities, and neonatal course (in intervals corresponding to the timed ultrasound to allow analysis by temporal sequence). A maternal postpartum interview - covering factors such as smoking, alcohol consumption, illnesses, reproductive history and antenatal was obtained from 78% of the study population. A major objective of this study is to use this extensive prenatal and perinatal data set to analyze the antecedents of neonatal brain hemorrhage. Follow-up (blind to birth history) consists of examinations of the infant's neurological status, hearing and vision, and assessments of cognition, language and behavior. An assessment of maternal psychopathology is obtained at two years, and maternal measures of stress and attitudes towards the infant are obtained both at birth and at two years. Our overall goal is to link prenatal and perinatal exposures to ultrasonographic findings, and to link these findings in turn to rates of neurodevelopmental impairment at age two.